Hi, I have been following this proposal with interest. I often have tried to tag hazards, and not found a good ways of doing it. We are now compiling a long list of hazards, including golf players crossing the road, but I see some basic aspects which are not being addressed (unless I missed something):
I would like to see signposted hazards completely separately tagged from hazards that the mapper perceives in a place, but which are not signed. Signed hazards should be mapped. - on nodes, if the extension of the hazard is point-like (example: dangerous railway crossing) - on ways, if the hazard exists along a highway (example: animals crossing zones) - (possibly) on areas, if the hazard is present in an area (example: landslides) In the case of signed hazards, I see two alternative ways of tagging the signing: - (only for nodes and ways highway segments) by adding source:xxx=sign like we do with speed limits - by mapping the relative signs as nodes Insertion of signposted hazards do not require any assessment of the presence of the hazard by the mapper. Signposted hazards are most often signalling dangers for vehicle drivers. Let's take the sign for hazard=cyclists (crossing), which warns clearly the vehicle drivers on the carriageway, that there could be cyclists crossing. There is normally no such warning on the crossing cyclists' path. There are exceptions of hazard warnings for both parties like a "cyclists sharing the road" sign, but that's the only one that comes to mind. Another aspect that should be defined: Are writings or pictograms on the road surface equivalent to vertical traffic signs? A completely different story are unsigned hazards with no signs on the ground, i.e. situations perceived as a hazard by the mapper. These are the tricky ones. I map cycling infrastructure, hence my examples come from that perspective. Examples: - foot-cycle crosswalks where there is a sign-posted speed limit of 30km/h, but where 90% of the cars pass with speeds far exceeding that value and making the place really dangerous - a two-way cycle path that is parallel to a main road and crosses a side road with a foot.bicycle crosswalk - car drivers entering the side road regularly overlook cyclists which ride in the same direction as they drive (to my knowledge the major cause of cyclists being killed in many countries. These in most cases in my part of the world have no danger signs. - And now consider the same situation with a row of trees between the cycle path and the main carriage way. - In my part of the world authorities put all kinds of bollards, arches, chicanes on cycleways (supposedly for the safety of cyclists, but in reality to keep car drivers from parking there). Many of these are grey metal objects that become invisible at night even if you have a good cycle light, as they have no reflective markers on them. The problem here is that the tagging will be based on my perceived version of ground truth. If I am a cyclist, I may be good at spotting hazards for cyclists. If I am a horse rider I will be good at mapping hazards for horse riders. Then we have also the asymmetric situations: e.g. car drivers are warned by a sign that there will be cyclists crossing, but the (bigger) hazard of cars hitting the cyclists on the same crossing is not signposted for cyclists. Volker On Sat, 5 Dec 2020 at 17:05, ael via Tagging <tagging@openstreetmap.org> wrote: > On Fri, Dec 04, 2020 at 09:48:27PM +0000, Paul Allen wrote: > > On Fri, 4 Dec 2020 at 19:56, Martin Koppenhoefer <dieterdre...@gmail.com > > > > wrote: > > > > Up until around ten years ago, a minor road went past the end of the > > runway at what passes for an airport. The planes could be so low on > > approach to the runway that there were traffic signals to prevent > > vehicles crossing the path of an aircraft. There were also signs > > warning of low-flying aircraft, which I referred to as "Give way > > to aircraft." > > Also at much larger airports. Brize Norton > (https://en.wikipedia.org/wiki/RAF_Brize_Norton), for example. > > https://www.openstreetmap.org/node/190194553 for one of the traffic > lights. > > ael > > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging >
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
